The ’90s saw many landmark events: the OJ Simpson trial, the impeachment of former President Bill Clinton, and perhaps much less seriously, the rise of frosted tips.

One slightly less earth-shattering event was the development of search-engine optimization, or SEO, which has come to define search-engine rankings—and in turn, what content internet users encounter—since not long after the first search engines did.

But SEO is now, in turn, being affected by another landmark event: the debut of generative AI. Since ChatGPT exploded last year, marketers are finding new ways to assist their clients’ sites’ search rankings using generative AI, whether that’s through content creation or internal linking protocols.

“There has been, for the past 15, 20, [or] more years, the ongoing, never-ending arms race between SEO optimizers and search engines,” George Strakhov, EMEA chief strategy officer at DDB and the founder of the company’s “hybrid creativity lab,” RANDDDB, said. “In content production, AI is just a very powerful weapon that was added to this arms race on the side of the SEO people.”

With that said, there’s a real possibility for things to get messy: “It does have the potential to be the final weapon that makes traditional search completely unusable,” Strakov told us.

Make it personal

For years, SEO marketers have worked to optimize the content on their clients’ websites in hopes that they will rank high in search results and more users will see them. Now, they’re leveraging generative AI tools to help do that work.

In some cases, generative AI is being used to help generate content for clients’ websites, which can help search-engine rankings, Joe Stoffel, SEO director at Marcel Digital, told Marketing Brew. Marcel Digital works to match the style and tone of content to clients by feeding client information to generative AI apps like ChatGPT, which it uses to help generate tailored content ideas. Marcel then leans on subject-matter experts to help fact-check and refine the AI-generated output, he said.

“We’re most heavily using it coming up with new topic ideas, but also doing then 60% or 70% of that heavy lifting on actual client content creation,” Stoffel told us.

Stoffel mentioned one client that saw a 555% YoY increase in traffic to a certain section of its site for the last three months after Marcel used generative AI to help retool it.

Stoffel said that before, the company “was not producing informative resource content” for its website, something, he said, that can help with organic search-engine rankings. “A big focus of ours over the last year has been working with them to do the heavy lifting. With this AI content, you get more of this resource information onto their website,” he said.

Generative AI can also help augment clients’ sites’ internal linking protocols, he said, which can also impact sites’ SEO rankings. Marcel uses the web crawler Screaming Frog along with ChatGPT to audit and map the internal linking of a site, which helps uncover potential linking deficiencies, Stoffel said; ChatGPT can then generate recommendations on where to place internal links.

Marcel isn’t the only agency using generative AI to help clients create online content that could help their search-engine rankings. DDB built a customized generative AI tool for an airline client to help its internal copywriters create travel articles, Strakhov said. The agency trained the tool on thousands of pieces of copywriting material and built a web-based interface that allows airline copywriters to specify details including word count and geographic destination; the tool spits out a first draft in the brand’s voice.

“Their department that is in charge of SEO is much more productive because they go ahead and go to the interface and say, ‘Give me the first draft for an article about the South China Sea,’” he said.

Traffic jam

Some agencies are using generative AI to more closely target customers as a way to drive traffic to client sites. Marcel uses ChatGPT to help generate website metadata as well as help draft calls-to-action that are tailored to the needs and pain points of certain “target personas,” Stoffel said.

At the marketing agency TriComB2B, staff are leveraging Google’s AI tools to help target customers with the goal of driving website traffic, according to Andrew Humphrey, the agency’s senior director of media strategy. He said that feeding analytics data from a client’s site into Google’s AI-powered ad tools can help target users who could be more likely to click on the client’s site and buy the client’s products, he told Marketing Brew.

“Not everybody who’s searching the same keyword, for instance, is going to be likely to buy,” Humphrey said. “I’m telling Google, ‘Yeah, this is the keyword I want to target, but also of them, if there’s 100 people also searching for that one keyword, who are the 10 of them most likely to make a purchase on my website?’”

Careful now

As agencies are finding diverse ways to employ generative AI in SEO marketing, they are also erecting different safety guardrails. Stoffel said that his team anonymizes customer information and client data before feeding it into ChatGPT Team, a workplace version of the popular tool. At TriComB2B, any unpublished client materials will not be fed into a generative AI program, Humphrey said. And at DDB, while generative AI can be used to create drafts, it can’t be used to generate final output, Strakhov told Marketing Brew.

“There is always a human looking at those drafts, editing them, and sometimes completely rewriting [or] regenerating them,” he said.